Auditor Settlements of Securities Class Actions

• Date: 01 March 2017
• Author: James J. Park
• DOI: http://doi.org/10.1111/jels.12144
• Published date: 01 March 2017

Auditor Settlements of Securities

Class Actions

James J. Park*

Some corporate law scholars have concluded that auditors do not have sufficient legal

incentive to detectsecurities fraud and should be governed by a strictliability standard. This

study assesses this argument by examining a dataset of 554 class actions alleging an

accounting restatement filed from 1996 through 2007. Because some but not all of these

restatement cases named an auditor defendant, it is possible to analyze whether variables

such as the liabilitystandard affect both the decisionto name an auditor defendant aswell as

the outcomeof the case. Despite the narrowingof auditor liability under Rule10b-5, auditors

are still often named as defendants and pay substantial settlements in Rule 10b-5 cases. A

more restrictive liability standard is associated with a modest reduction in the rate at which

auditors are named as defendants and the rate at which auditor cases end in settlement.If a

Rule 10b-5 case against an auditor is strong enough to result in a settlement, the legal

standard does not affect the size of the settlement. The auditor’spayment is correlated with

nonlegal factorssuch as whether the issuer is bankrupt and the issuer’smarket capitalization.

These results arebest explained by the tendency of judgesto read narrow liability provisions

broadly in cases wherethe size and impact of the alleged fraudare significant. The evidence

thus does not support the conclusion that a strict liability standard is necessary to generate

sufficientincentives for auditors to detect substantial frauds.

I. Introduction

Because of the critical role of a company’s financial statements in valuing its securities,

auditors serve as a critical gatekeeper to the capital markets. The federal securities laws

thus make auditors liable to purchasers of securities in certain circumstances when a

company commits securities fraud under the auditor’s watch. From the late 1990s to the

mid-2000s, an epidemic of accounting fraud at major public companies led many com-

mentators to question why auditors had not detected such misconduct.

One explanation was that Congress and the courts had shielded auditors from pri-

vate litigation under the securities laws, reducing the incentive of auditors to carefully

*Professor of Law, UCLA School of Law. Address correspondence to James J. Park, 385 Charles E. Young Dr. E.,

Los Angeles, CA 90095; email: james.park@law.ucla.edu.

Thank you to participants at the American Law and Economics Association Annual Meeting (2015), the Secu-

rities Regulation Section, American Association of Law Schools Annual Meeting (2015), the Conference on

Empirical Legal Studies (2015), and the Corporate and Securities Litigation Workshop (2015) for helpful com-

ments. Thank you to Joseph Doherty for statistical consultation and Matthew Brodsky for research assistance.

169

Journal of Empirical Legal Studies

Volume 14, Issue 1, 169–198, March 2017

audit public companies. First, in its 1994 Central Bank v. First Interstate Bank decision,

1

the U.S. Supreme Court ruled that secondary actors such as auditors could not be liable

for aiding and abetting under SEC Rule 10b-5, the most widely used cause of action for

securities fraud. Instead, the plaintiff must show that the auditor was a primary violator

of the law, a difficult task because an auditor is rarely the main architect of a fraud. Sec-

ond, in 1995, the Private Securities Litigation Reform Act (PSLRA), which was heavily

supported by the auditing industry, made it more difficult to successfully bring claims

under Rule 10b-5. Among other reforms, after the PSLRA, plaintiffs have a greater bur-

den to establish a strong inference of fraudulent intent at an early stage of the lawsuit.

The courts have made establishing such intent especially difficult with respect to an

auditor defendant.

These developments prompted a leading commentator to note that “auditors ...

have been well insulated against securities fraud liability” (Coffee 2006:1550). Writing

soon after the Enron and WorldCom scandals, several corporate law scholars indepen-

dently concluded that auditors should be governed by a strict liability standard (Partnoy

2001; Hamdani 2003; Coffee 2004). At the time of these proposals, there was little signif-

icant data on securities class actions filed against auditors after the liability shielding

efforts of the mid-1990s.

2

A later strand of the literature has documented that auditors continue to pay sig-

nificant amounts to resolve securities class actions. In studying the possibility that an

auditor could be subject to catastrophic liability, scholars have noted that auditors con-

tinue to be defendants in securities class actions and pay significant settlements (Talley

2006; Donelson 2013). Another study finds 144 auditor settlements in cases filed from

1996 to 2005, and that such settlements are more likely in securities class actions involv-

ing an accounting restatement (Donelson & Prentice 2012).

Though some scholars have used theory to analyze the legal standard governing

auditors, and others have studied cases imposing significant auditor liability, no study has

closely examined how restrictions on the scope of liability affect the resolution of securities

class actions against auditors. This study analyzes a dataset of 554 securities class actions

involving restatements filed from 1996 to 2007, and finds that some liability restrictions are

associated with a modest reduction in the rate at which auditors are named as defendants

and the rate at which auditor defendants contribute to a settlement. However, once a case

settles, the size of that settlement is not affected by the legal standard.

In coming to these conclusions, this study exploits two differences in the way secu-

rities class actions are resolved in assessing the impact of more restrictive auditor liability

standards. First, with respect to the elimination of aiding and abetting liability, some cir-

cuits retained a test with a close resemblance to the old aiding and abetting test. Most

notably, the Ninth Circuit considered an auditor to be a primary participant when it

1

511 U.S. 164 (1994).

2

For example, one of the articles lists six examples of auditor settlements during this time period (Coffee

2004:341), and another article noted that “settlement value of cases against auditors has gone way down” (Coffee

2002:1410).

170 Park